[sachu]

Quote:

Sachu, 
 
BTW, do you know the bias value for the production LL? I think Kevin posted the values for Stax and his amps before somewhere but I've lost track of that.
 
Arnaud

 
 
Hi Arnaud,
 
First of all, thanks for helping in acquiring the sample TKD pot for me.
 
As for your question, you should pose it to Alex on his sponsored thread. He is surely going to be forthcoming on it.
 
I am not involved with Cavalli Audio in any capacity at all anymore and am not privy to his design decisions anymore. I only used to do some prototyping work now and then for him, but since January have stopped even that.

 

[sachu]

and you must be referring to this thread
 
http://www.head-fi.org/t/582518/electrostatic-amplifiers-voltage-ratings

 

[arnaud]

sachu, yes exactly... Thanks!

 

[kevin gilmore]

Those are peak to peak voltage ratings.
The output bias stage current ( I.e class a power)
Is something completely different. Since the load is a capacitor
Calculations are a bit different and the power can translate into
Slew rate.

The total output power of the t2 for example is 18 watts yet anyone
That has touched the heatsinks on one knows how warm it gets.
( watts per channel)

The bhse and kgsshv are in the same range.

The srm323 and srm727 are much lower.

For many amps I can calculate the power.

40 watts of class A power total is a significantly high number

 

[arnaud]

Based on Kevin's thread and Alex's reply in the sponsor thread, the LL sits in between the Stax mps and BHSE in terms of output stage bias current:

- Stax 727: 350V rails / 1400V PP / 6.6mA
- KGSS: 350V rails / 1400V PP / 8.3mA
- Cavalli Audio LL: 400V rails / 1600V PP / 13mA
- BHSE: 400V rails / 1600V PP / 18mA
- DIY T2: 500V rails / 2000V PP / 14mA

I did not list the KGSSHV as it can come in different flavors as far as voltage swing and class a bias...

Also, wouldn't be straight forward to add how this translates in terms of max power at like 1kHz into SR007 / 009 headphones since their capacitance is known (and you're mostly driving a capacitive load)? Isn't the impedance like 1/( j C omega )?

Edit: corrected BHSE info.

 

[gilency]

Per Kevin Gilmore's information:
 
[size=medium]2000 volts peak to peak stator to stator  (+/-500 power supplies)[/size]
[size=medium]KGSSHV (ixys parts)[/size]
 
[size=medium]1800 volts peak to peak stator to stator (+/-450 power supplies)[/size]
[size=medium]KGSSHV (sanyo parts)[/size]
 
5.5ma x 4  (you can turn it up if you have lots of heatsink)

 

[seacard]

Quote:

Based on Kevin's thread and Alex's reply in the sponsor thread, the LL sits in between the Stax mps and BHSE in terms of output stage bias current:
- Stax 727: 350V rails / 1400V PP / 6.6mA
- KGSS: 350V rails / 1400V PP / 8.3mA
- Cavalli Audio LL: 400V rails / 1600V PP / 13mA
- DIY T2: 500V rails / 2000V PP / 14mA
- BHSE: 500V rails / 2000V PP / 18mA
I did not list the KGSSHV as it can come in different flavors as far as voltage swing and class a bias...
Also, wouldn't be straight forward to add how this translates in terms of max power at like 1kHz into SR007 / 009 headphones since their capacitance is known (and you're mostly driving a capacitive load)? Isn't the impedance like 1/( j C omega )?

 
Anybody know where the Electra will come in?

 

[kevin gilmore]

So here is an example.
 
From stax website on sr009
110 pf including cable
at 10khz the impedance is 145kohm
 
So lets pick a number, say 1120 volts peak to peak stator to stator
that is 400 vrms.
 
400 vrms into 145 kohm at 10khz is  1.1 watt
at 20 khz it is 2.2 watts
 
at 1khz its 100 milli watts.
 
This is however a theoretical power if the load was actually resistive
which it clearly is not.
 
Of course actually listening to 20khz at that level is bound to damage
your hearing.
 
Craig has said that the electra is going to be about 1800 vppss
 
Also the BHSE is 400 volt rails.

 

[YOONG 2112]

My T1 has still the original Japanese "Golden Arrows for Stax" tubes in them. the sound is " ok." since I have no other similar amp to make a direct comparison.The set would be at least 20 years old by now .! Is it time to change the tubes? Are tubes "used by" time dependent or "usage hours" dependent. ? I have no idea how many hours it has clocked as I bought it 2nd hand.Thus If deemed necessary, I have  3 different pairs of  6CG7/6FQ7 of different brands for replacement ie RCA clear top, Sylvania and Raytheon ( these are NOT NOS but all used and tested to work ok). Which of the three would best suit the T1? Any feed back on the sound characteristics of these three?
Also the T1 has a circuit board with a number PB-201 printed, Can anyone please help with instruction that can guide me on how to set the correct bias ? Much appreciated.!

 

[Michgelsen]

Instructions to adjust the bias of the SRM-T1 are found here: http://www.head-fi.org/t/324155/biasing-stax-srm-t1
 
If you have new tubes and are going to rebias the amp anyway, why not try it with new tubes? You'll have to open the amp either way. I've read good things about the the RCA Clear Tops in Stax amps, but have no experience with them. When I had my SRM-T1 I had some nos NEC tubes installed which were good as well.
 
Have you ever switched tubes before? In case you haven't, it must be noted that pulling out the tubes must be done very carefully, while gently moving them around a little bit but never twisting them. This way they slowly come out of their sockets. It can feel like they're almost stuck in there. Pulling too hard at once may damage the PCB. Putting in the new ones can be done in the same way.

 

[di_andrei]

Quote:

 
They changed the driver or rather the diaphragm which did alter the sound quite a bit.  It now uses the same diaphragm as the SR-009 (which Stax won't disclose anything about).  The SZ2 version used pretty much the same driver as the Mk1 but they might have altered it a bit to get a brighter sound.  I certainly never got them to match a Mk1 100% but I got damn close.  

I would really like to see a guide to the versions of 007, preferably with photos - a lot of the older sets out there have had their headbands replaced so the serial numbers are gone.

 

[arnaud]

Kevin, many thanks for the additionalrrv clarification. I corrected the BHSE spec.

While we're at Electronic Circuits 101, and if you (or anyone with the knowledge)  don't object to teach some to the fools like myself:
1. Is the dissipated power at idle related to the bias current on the output stage (it would seem like as the larger amps require large heat sinks and the capacitive load does not dissipate anything)?
2. Be it tubes or transistors, they both dissipate a lot of energy into heat, even when they're not loading / driving anything, and especially if they're heavily biased into class A?
3. How to derive the slew rate from a given voltage swing target (which I guess I can estimate from the sensitivity and desired SPL) / current bias / load capacitance?
4. Conversely, how to define the necessary slew rate for a given load / desired voltage swing? 

 

[kevin gilmore]

Quote:

- Stax 727: 350V rails / 1400V PP / 6.6mA
- KGSS: 350V rails / 1400V PP / 8.3mA
- Cavalli Audio LL: 400V rails / 1600V PP / 13mA
- BHSE: 400V rails / 1600V PP / 18mA
- DIY T2: 500V rails / 2000V PP / 14mA
 

First a comment on this one. 4 of the above are output stage current per driver. multiply by 2 to get the output
stage current per channel.  The cavalli number is likely the current per channel because otherwise the amp
would be doing 40 watts of heat inside that box, plus the power supply power and that  seems unlikely.
 
1) the heat is directly related to the output stage current, where most of the power goes. Some additional heat
comes from the driver stage and input stage.
 
2) same as 1 above. a class A device idles at a specific power level. Music does not change the power level.
The maximum efficiency is something like 40%, but for electrostatics, much much lower.
 
3 and 4)  Now this gets complicated, and more power does not necessarily indicate a faster slew rate, but
usually does. To calculate minimum slew rates, many people assume a sine wave at 20khz, know what
the maximum peak to peak voltage is, and calculate what you would need to get there. Its the huge
voltage swings that are at issue. A dynamic amp can get away with 5 volts per microsecond. An electrostatic
amp needs at least 50 volts per microsecond, and 200 volts per microsecond is way better, only the T2
does that.

 

[arnaud]

Thanks again Kevin. Good point about the current per driver vs channel, I will check.

Otherwise, reading again your other thread / your exchanges with Frank Cooter, it would seem that:
- Slew Rate ( dV/dt ) assuming a sine wave is: SR = Vp x 2pif (typically in V/micro second).
- Minimum target for slew rate is (500 Vp) SR = 60V/micro sec at 20kHz
- For a capacitive load, the instanteneous current and voltage are related by i(t) = C x dV/dt = C x SR
- The sr009 has 110 picoF capacitance so that translates into minimum current of 6.6mA (per driver)
- A capacitive load has an impedance as Z=1/(j omega C).
- The sr007mk2 has a 150kOhm impedance at 10kHz, which translates into 106 picoF and 6.3mA (per driver)

Assuming the above is correct, my last questions:
1 Is it that the output stage can only deliver as much current as its bias before going non linear?
2. How to estimate the dissipated power (from the output stage) from the max voltage and current bias?
3. Can we estimate the amp's slew rate from other specs or it has to be measured?

 

[runeight]

Just a small correction. The 13mA is per CCS not per channel.